Tuesday 29 July 2014

Avoiding death: a few ideas

The Deary gang are at it again, this time looking at US health data to show that reaction times are related to mortality. Naturally, I did not wish to think about this paper whilst lying supine on the beach, but now reality presses in on me again, mixed with post-holiday good resolutions regarding exercise and restraint in all pleasurable matters. Although the subject of mortality is dispiriting, imagine my joy to find that their first table consisted of means and t tests, the sorts of statistics I can understand.

Reaction Time and Mortality from the Major Causes of Death: The NHANES-III Study. Gareth Hagger-Johnson, Ian J. Deary, Carolyn A. Davies, Alexander Weiss, G. David Batty


In a meta-analysis comprising 16 studies of over one million participants, a 1
standard deviation increase in cognitive ability in childhood was associated with 24% lower risk of mortality. Reaction time and cognitive ability may both predict mortality risk because they both measure important aspects of neuropsychological functioning or reflect the integrity of one or more bodily systems. However, reaction time is also seen to explain the IQ-mortality association suggesting that it may mediate the association between more complex cognitive processes and mortality. The analytic sample comprised 5,134 adults (2,342 men) aged 20 to 59 with data on reaction time and who were followed for mortality for 15 years (378 deaths).

So, here in Table 1 are a few tips to keep you alive: better to be a woman than a man (though chopping off reproductive organs and taking hormone tablets will not help you, however appealing the ploy may be as a conversational gambit); probably best not to belong to an ethnic minority (American blacks and Mexicans); possibly worth completing your schooling; certainly worth while not being in a lowly occupation and on a low income; and certainly unwise to be a regular smoker and drinker (though no risk in being overweight or obese, or eating saturated fats). Of course, not all of these are under your control, because although most of us aspire to high status, high income jobs, others want them as well, leading to competition and consequently to social stratification. Blood pressure is a risk factor (probably a consequence of other factors) but serum cholesterol is not.



Into the familiar list of health warnings (notice that in this sample many apparent risks do not prove significant) enters the one behavioural measure, not a full IQ test but a lowly simple reaction time test, which proves to be as significant as the other most significant factors.

No variables attenuated the associations fully, suggesting that the association between simple reaction time and mortality is independent of socio-demographic, socio-economic, health behaviors and CVD risk factors.

It would appear that reaction time gives us an indication of system integrity, a hypothetical construct which determines physical health and is associated with having a sharp and alert mind. Just in case you were considering it, it is not possible to improve simple reaction time by continual practice. Ian Deary tried it by doing the test every morning for about three months, and found that such improvements as were possible were achieved on the second day, and with absolutely nothing thereafter.

So, if you are lying on a  beach, there is no need to practice fast reactions, but you might consider not smoking and drinking and eating so much, once the holiday is over.


  1. What I find interesting is that controlling for "health behaviors" tends to partially (in some studies completely) attenuate the relationship between simple reaction time and health. While at the same time, these factors have no such effect on the relationship between youth IQ and health.

    Perhaps this indicates that simple reaction time is itself affected by declining health. Maybe we should find a study that compares health outcomes with reaction time as measured in youth or early adulthood.

  2. Thanks. I should have made it clear that Deary excludes any reaction time data if they are collected within 5 years of the person's eventual death, this being a prudent exclusion to discard any close to death slowing up of responses. As regards "health behaviours" as you have pointed out, these are g loaded behaviours, so they are different ways of accounting for the factors which make up the nexus of intelligence, health and reaction time. There is something there: that is the exciting finding.

  3. I've always loved the idea that life is an IQ test, and the positive correlations between IQ and income, occupational status, and life span are just three examples of that. In other words, perhaps high IQ folks have longer and more successful lives simply because they avoid making mistakes and make more wise decisions. After all a biologist might define intelligence as the ability to adapt; to take whatever situation you're in and turn it around to your advantage. Obviously, death is highly disadvantageous so high IQ people should be good at avoiding it (though it's usually beyond one's control).

    So it would be interesting to look at studies correlating IQ with preventable death vs correlating IQ with unpreventable death to determine whether the correlation is mediated by actual intelligent behavior or whether it's simply a function of IQ and life span both being caused by a third variable (general biological quality). Of course such a dichotomy between deaths is highly simplistic, insensitive and offensive, but I do wonder if people who die young because they made a mistake (i.e. driving fast in a snow storm) are less intelligent than people who die young because they're just naturally unhealthy.

  4. Presume you have seen this excellent paper, which set out the issues many years ago. http://www.udel.edu/educ/gottfredson/reprints/2004currentdirections.pdf
    There is still a case to be made that clever behaviour leads to a safer and longer life (so long as you avoid private planes) but the RT data shows that part of the benefit seems to be in addition or in parallel to IQ, and not all of it because of IQ.